Compliance at GoCrazyAI

1AI Transparency & Content Marking

Everything created on GoCrazyAI is AI-generated, and we make that clear in three layers — to the people using the platform, to the people who later see the content, and to software that checks files automatically. This implements the transparency obligations of Article 50 of the EU AI Act (Regulation 2024/1689).

• Machine-readable marking (Art. 50(2)). Every image, video, and audio file generated on the platform from August 2026 onward carries an embedded, machine-readable marker identifying it as AI-generated: the industry-standard IPTC Digital Source Type value trainedAlgorithmicMedia, written directly into the file’s metadata (XMP for images and video, ID3 for audio) at the moment the file is saved. This is the same vocabulary that major platforms and search engines read to recognize and label AI-generated media. The marker is deliberately anonymous: it identifies the content as AI-generated and contains no personal data about the user who created it.
• Marking is reserved for AI output. Media that users upload as source material — their own photos, videos, or voice recordings — is never given the AI marker. Only content produced by our generation pipelines is marked, so the marker stays meaningful.
• Visible disclosure (Art. 50(4)). Pages and tools that produce lip-sync, avatar, or other likeness-altering media carry a visible notice that outputs are artificially generated or manipulated, and our Terms require users to clearly label such content as AI-generated whenever they publish it, and forbid removing labels or provenance metadata from generated files.
• Interaction disclosure (Art. 50(1)). Every conversational AI surface — AI Characters and AI chat — displays a persistent notice that responses are AI-generated and that the user is not talking to a real person.
• Text outputs. Plain conversational text has no metadata container and text watermarks do not survive copy-paste, so — consistent with the “technically feasible” qualification in Art. 50(2) — text relies on the interface-level disclosure above rather than embedded marking.
• Next step. We are evaluating cryptographically signed provenance (C2PA Content Credentials) on top of the current marking, which would make the AI-generated attestation verifiable by third parties.

2Adults-Only Platform

• GoCrazyAI is available exclusively to users who are 18 years of age or older. Age eligibility is an express condition of our Terms and is affirmed at account creation.
• Our generation pipelines apply dedicated screening designed to prevent any depiction of minors in inappropriate contexts, and imagery policies across the product (for example, music cover art) exclude depictions of children entirely.

3Content Moderation

Moderation runs before generation, not after the fact, and decisions are logged.

• Prompt screening. Text prompts are screened before any generation starts. Requests that violate our prohibited-content rules are refused and the generation never runs.
• Image screening. Uploaded reference images pass automated checks — including checks specifically protecting minors — before they can be used in generation. Screening events are logged for audit and enforcement.
• Enforcement. Violations can lead to content removal, credit forfeiture, and account termination as described in our Terms.

4Likeness, Voice & Consent

• Users may only work with photos, videos, and voice recordings of themselves or people who have given consent. Impersonating real people, misleading likeness use, and non-consensual content are prohibited uses under our Terms.
• Voice cloning requires the user to confirm they own the voice sample or hold the speaker’s consent before a clone is created.
• Full detail lives in Trust & Safety, section “Consent, Likeness & Identity Protection.”

5Image Removal Requests

• Anyone who believes an intimate or manipulated image of them was created or shared without consent can file a removal request at gocrazyai.com/report-content. We review and act on valid requests within 48 hours, consistent with the US TAKE IT DOWN Act, and this removal channel is linked from every page footer.
• Reports of any other policy violation can be sent to go@gocrazyai.com or raised through in-app support.

6Consumer Protection

• Right of withdrawal. EU/EEA consumers can exercise their 14-day right of withdrawal through a dedicated, always-accessible withdrawal page (EU Directive 2023/2673). Where digital content is delivered immediately, consent and waiver are collected explicitly, as the Directive requires.
• Clear recurring billing. Subscriptions state the renewal price and interval before purchase and require an explicit consent checkbox for recurring charges. Cancellation is self-service and takes no more than a few clicks from account settings.
• Refunds. Our Refund Policy is published and applied consistently; failed generations are automatically refunded in credits without the user having to ask.

7Privacy & Data Control

• Our Privacy Policy describes what we collect and why. We do not sell personal data.
• Deletion on request. Users can delete their account and associated data at any time from account settings; deletion removes personal data and stored creations from our systems.
• Marketing email includes one-click unsubscribe, honored immediately.

8Payment Security & Monitoring

• All card payments are processed by PCI DSS–compliant payment providers. GoCrazyAI never stores card numbers on its own infrastructure.
• The platform’s public content and payment surfaces are independently and continuously monitored by Regra for payment-network and AI-regulation compliance — the verification badge in our footer links to the live status.
• Purchases and credit movements are recorded in an internal ledger so that billing disputes can be resolved against an auditable record.